PROJECT SUMMARY The National Institute on Aging and Alzheimer’s Association recently proposed the ATN framework to systematically classify Alzheimer Disease (AD) in research studies and integrate standardized biomarkers with clinical assessment. The framework is based on binary designations of the presence or absence of amyloid (A), tau (T) and neurodegenerative (N) pathology, wherein AD is defined as A+/T+/N±. This biomarker discretization may simplify AD diagnosis and management but does not enable a more quantitative dissociation of T and N beyond the statement of T+/N±. We are interested in studying neuronal responses to T pathology, including susceptibility (T<N) and resilience (T>N). N has been defined to include structural volume loss (NS) on magnetic resonance imaging (MRI) and lower neuronal metabolism (NM) on positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG). There has not been substantial investigation comparing T with NM, which may necessitate a more nuanced, comparison beyond the binary descriptions. Here, we will utilize the AD Neuroimaging Initiative (ADNI) to investigate the match and mismatch of concomitant T and NM with 18F-flortaucipir and FDG PET. We aim to (1) define “T/NM mismatch” measures derived from several methods including region-of-interest clustering, voxelwise thresholding and deep learning, (2) evaluate the relationships between T/NM mismatch with clinical factors (including cross-sectional and longitudinal cognition, prognosis and progression of T pathology) and (3) assess relationships between T/NM mismatch additional measures of N (including structural NS). Whether in the presence or absence of disease-modifying AD therapy, our in vivo molecular neuroimaging strategy may empower a more comprehensive understanding of the localization and sequence of pathological events leading to AD and unique biological responses to T, including functional resilience.